Passive crowd sourcing for technology prediction

Abstract

Technology prediction methods use various types of information to make systematic forecasts about technological innovations. Forecasting approaches vary, including quantitative (such as patent analysis) and qualitative methods (such as expert elicitation using the Delphi method). We discuss a new method and system for predicting technology futures by harnessing the predictive information made available by society in open sources. Our approach automatically discovers futurelooking temporal phrases associated with technology topics and presents predictions deemed significant using the G2 statistic. Here, we evaluate the phrase discovery component using a dataset of 782 technology forecast statements. We hope to demonstrate that’passive crowd-sourcing’ may be a meaningful source of technology-related predictive intelligence.